Human Microbiome: Methods, Clinical Applications, and Recent Advances

A patient has had Clostridioides difficile diarrhea three times this year. Each course of antibiotics clears it for a few weeks, then it comes back. On the fourth episode, the team does something that would have sounded absurd a generation ago. Instead of another antibiotic, they give the patient stool from a healthy donor, screened and processed, delivered as a capsule the patient swallows. Within days the diarrhea stops, and this time it stays gone.

The antibiotics were attacking the wrong problem. They kept killing C. difficile, but they also kept killing the hundreds of other gut bacteria that normally hold C. difficile in check. The cure was not a specific drug. It was restoring the community. That community is the human microbiome, and learning to read it and repair it is one of the fastest-moving areas in medicine today.

What the microbiome is, and why "normal flora" was only half the picture

For most of the history of microbiology, we studied the microbes on and in the body by growing them. We swabbed a site, streaked a plate, waited for colonies, and named what grew. This gave us the idea of normal flora: the resident bacteria, fungi, and other microbes that live on the skin, in the mouth, in the gut, and elsewhere without causing disease. If you want the organism-by-organism roster of what lives where, that is covered in detail in our article on the normal flora of the human body.

Culture had a hidden problem. Most microbes on the body will not grow on a standard plate. They need partners, exact gas mixtures, or growth signals we cannot easily simulate in the lab. So the plate showed us the few organisms that grow easily and hid the rest. For decades we were describing a community while seeing only a small part of it.

Two ideas fixed this, and they are the reason "microbiome" is now the preferred word.

1. First, a shift in method. Instead of growing microbes, we now read their DNA directly from a sample. We do not need the organism to grow. We only need its genetic material. This one change revealed communities far larger and more diverse than culture ever showed.
2. Second, a shift in the question. "Normal flora" asks who is present. "Microbiome" asks a bigger question: who is present, what genes do they carry, and what are those genes doing. The distinction matters because two people can carry different species that perform the same job. What the community does can matter more than exactly which species are doing it.

So the two words are not the same, and the difference is worth fixing in your mind early:

• Microbiota: the organisms themselves, the actual species living at a body site.
• Microbiome: the organisms plus their collective genes and what those genes produce. Some sources use "microbiome" for the genes alone, but the broad usage above is the practical one for clinical work.

The human microbiome is genuinely large. An adult carries roughly the same number of microbial cells as human cells, and the collective microbial genes outnumber human genes many times over. This is why the microbiome is sometimes described as a "forgotten organ." It performs digestive, protective, and signaling work that our own cells cannot.

The major body-site communities

The microbiome is not one thing. Each body site is a different habitat, with its own temperature, moisture, oxygen level, and food supply, so each carries a distinct community doing distinct work. The point here is to understand what each community does and why that matters clinically.

Gut microbiome. The largest and most studied community by far, concentrated in the colon. It ferments the dietary fiber we cannot digest and releases short-chain fatty acids that feed the cells lining the colon and influence metabolism throughout the body. It makes vitamin K and several B vitamins. It trains the immune system, and it holds pathogens in check by occupying space and consuming nutrients. When people say "the microbiome" without a qualifier, they usually mean this one.

Oral microbiome. The mouth is not one habitat but several: teeth, tongue, gum line, and saliva each carry different communities. This community is directly tied to two of the most common diseases in the world, dental caries and periodontal (gum) disease, both of which are now understood as shifts in the oral community rather than the work of a single germ.

Skin microbiome. Adapted to a dry, salty, nutrient-poor surface, and denser in warm moist areas like the armpit and groin. It defends the skin by competing with pathogens and by signaling to the skin's immune cells. The organism-level detail, and how to read "skin flora" on a culture report, is covered in our article on the normal flora of the skin.

Vaginal microbiome. Unusual because a healthy community is dominated by a single group, Lactobacillus, which produces lactic acid and keeps the local pH low and hostile to many pathogens. When this dominance is lost, the community diversifies in a way that is actually a sign of disease. This is the reverse of the gut, where high diversity generally signals health. The direction that counts as "healthy" depends on the site.

Airway microbiome. For a long time the lower airway was taught as sterile. Sequencing showed that the healthy lung carries a sparse but real community, seeded mainly by tiny amounts of inhaled and aspirated material from the mouth. It is low in biomass, which makes it technically hard to study and easy to contaminate during sampling, so read airway microbiome claims with more caution than gut claims.

One theme runs through all five. A healthy microbiome is not defined by the same rule everywhere. High diversity is healthy in the gut but the opposite in the vagina. "Balanced for that site" is the real standard.

Dysbiosis: when the community stops working

Dysbiosis means a disturbance of the microbial community that is linked to disease. Dysbiosis is not simply "some bad bacteria appeared." It usually means one or more of the following:

• Loss of diversity. Fewer types of organisms, so the community is less stable and less able to resist an invader. This is the classic pattern after broad-spectrum antibiotics.
• Loss of key members. A few organisms do jobs the rest cannot, such as producing butyrate, a short-chain fatty acid that feeds colon cells and calms inflammation. Lose those members and the function is lost even if total numbers look normal.
• Overgrowth of a minor member. One organism that was previously kept small expands and dominates. C. difficile after antibiotics
• A shift in what the community does. The species list can look almost normal while the community's output, its metabolites and signals, has changed. This is why reading genes and products (the microbiome) tells you more than reading the species list (the microbiota) alone.

The clinical importance of dysbiosis is that it is increasingly seen not just as a marker of disease but, in some conditions, as part of the cause. That is the idea that makes microbiome-based treatment possible: if a disturbed community drives disease, then restoring the community may treat it.

How we actually study the microbiome

This is the part that separates modern microbiome science from classic flora work, and it is worth understanding at the level of what each method can and cannot tell you.

Culture. Grow the organism, then study it. Still essential for isolating a specific pathogen, testing antibiotic susceptibility, and confirming that an organism is alive. Its limit is coverage: it misses the majority of the community that will not grow on standard media.

16S rRNA gene sequencing. The workhorse of early microbiome studies and still widely used because it is cheap. Every bacterium carries a gene called 16S rRNA. Part of this gene is nearly identical in all bacteria, and part of it varies from group to group. We copy and read that variable part, then match it to a database to identify who is present. Its strength is a fast, affordable census of bacteria.

Its limits are two: it usually identifies organisms only to the genus level, not the exact species or strain, and it tells you who is there but not what genes they carry or what they are doing. It also reads bacteria and archaea, not fungi or viruses.

Shotgun metagenomics. Instead of reading one marker gene, we chop up and read all the DNA in the sample. This is a major step up. It identifies organisms to species and often strain level, it captures fungi and viruses as well as bacteria, and, because it reads every gene, it tells you what the community is capable of doing, including which antibiotic resistance genes are present. Its cost is higher, and it needs more computing power to analyze.

Metatranscriptomics. Metagenomics reads the DNA, so it shows what the community could do. Metatranscriptomics reads the RNA, which shows which genes are actually switched on at that moment. This is the difference between owning a tool and using it. A resistance gene that is present in the DNA but never expressed behaves differently from one that is active. RNA is fragile and changes fast, which makes this method powerful but technically demanding.

Metabolomics. This one skips the microbes and measures their products directly: the short-chain fatty acids, bile acid modifications, and other small molecules the community releases. Since these products are often how the microbiome affects the body, measuring them can be the most direct read of function. It is usually combined with the sequencing methods above rather than used alone.

Put simply, there is a ladder from who is there to what they are doing: culture and 16S answer "who," metagenomics answers "who and what they can do," and metatranscriptomics plus metabolomics answer "what they are actually doing right now." More function means more cost and more technical difficulty.

A recurring trap sits underneath all of these methods. Sequencing detects DNA whether the organism is alive or dead, so a sequencing result is not proof of a living, active community. And in low-biomass samples such as the airway, or blood, or the long-debated question of whether the healthy womb has its own microbiome, the tiny amount of real microbial DNA can be swamped by contamination from reagents and the environment. Much of the early claim that the fetus is colonized before birth is now attributed to exactly this kind of contamination. When a microbiome finding comes from a site with very few microbes, ask how contamination was ruled out before you trust it.

Microbiome and disease

The reason to learn all of this is that the microbiome is now tied, with varying strength of evidence, to a wide range of disease. The evidence is strongest where we can both explain a mechanism and change the outcome by changing the microbiome. It is weakest where we only have an association.

Recurrent C. difficile infection (strongest evidence). Antibiotics wipe out the gut community that normally holds C. difficile down. With its competitors gone, C. difficile overgrows and releases toxins that inflame the colon. The proof that the microbiome is central here is direct: restoring a healthy donor community cures the infection when antibiotics alone fail.

Inflammatory bowel disease (IBD). Crohn disease and ulcerative colitis show reduced diversity and a loss of anti-inflammatory, butyrate-producing organisms. The likely mechanism is a feedback loop: a disturbed community drives inflammation, and inflammation further disturbs the community. Whether dysbiosis starts the disease or results from it is still debated.

Obesity and metabolic disease. Gut communities differ between people with and without obesity, and they influence how many calories are extracted from food and how the body handles blood sugar and fat. Animal studies show that transferring a microbiome can transfer metabolic traits. In humans the effect is real but smaller and more variable, so the microbiome is one contributor among many.

The gut-brain axis. The gut and brain communicate through the vagus nerve, through immune signals, and through microbial metabolites that reach the bloodstream. Gut communities differ in some people with depression, anxiety, and Parkinson disease. This is one of the most exciting and most overhyped areas at once: the communication is real and the associations are real, but a difference in the microbiome is not the same as a cause of the brain condition.

Drug metabolism. Gut bacteria chemically modify many drugs, sometimes activating them, sometimes inactivating them, sometimes producing toxic byproducts. This means two patients on the same dose can get different effects partly because of their microbiomes. It is an early but clinically important idea: the community is a hidden variable in how medicines work.

Across all of these, hold one line in mind. An association (the microbiome differs in a disease) is weaker evidence than a mechanism (we can explain how it contributes) which is weaker than a demonstrated intervention (changing the microbiome changes the disease). Recurrent C. difficile clears all three bars. Most other conditions clear only the first or second so far.

Applied and clinical translation: therapies, diagnostics, and products

This is where microbiome science becomes something you will prescribe, interpret, or counsel patients about. It is also where a large industry has grown up, some of it evidence-based and some of it not, so judgment matters as much as facts.

Fecal microbiota transplantation (FMT). Transferring processed stool from a healthy, screened donor into a patient to rebuild the gut community. Its established use is recurrent C. difficile infection that has not responded to antibiotics, where cure rates are high.

Donor stool is screened rigorously, because transplanting a community means transplanting whatever pathogens the donor carried, and serious infections have been transmitted when screening failed. FMT for conditions beyond C. difficile, such as IBD or metabolic disease, is still experimental. Read more about this in fecal transplant article.

Approved live biotherapeutic products. Until recently, FMT was a non-standardized procedure that varied between centers. The US FDA has now approved two standardized, manufactured microbiome-based products for preventing recurrent C. difficile, both derived from screened human donor stool:

• Rebyota (fecal microbiota, live-jslm), approved November 2022, given as an enema at a healthcare facility.
• Vowst (fecal microbiota spores, live-brpk), approved April 2023, the first such product taken by mouth as a capsule.

These approvals matter beyond C. difficile. They mark the point where "give the patient a healthy community" moved from an improvised procedure to a regulated, prescribable medicine. Expect more such products, for more conditions, over your career.

Probiotics. Live microorganisms, usually specific strains of Lactobacillus or Bifidobacterium, taken to support a healthy community. The key thing to teach patients is that "probiotic" is not one thing. Effects are strain-specific and condition-specific: a strain that helps one condition may do nothing for another, and most supermarket products have weak evidence for the broad claims on the label.

Probiotics are not the same as the approved live biotherapeutic products above, which are regulated as medicines and tested in trials. For the general concept and its evidence, see our article on probiotics and their benefits.

Microbiome-based diagnostics. Sequencing a patient's community to detect a disease signature or predict a response to treatment. This is an active research area with real promise, especially in colorectal cancer screening and in predicting which patients respond to certain cancer immunotherapies. Most of these tests are not yet routine clinical tools.

Direct-to-consumer microbiome testing kits. A patient can now mail a stool sample to a company and receive a report on their gut "health," often with diet or supplement recommendations. This is the part to counsel patients on carefully. The sequencing may be real, but the interpretation usually runs far ahead of the science. We do not yet have a validated definition of a single "healthy" microbiome to compare an individual against, results vary between companies and even between samples, and the recommendations are often generic. These kits can be interesting, but a report calling a bacterium "good" or "bad" is making a claim the science cannot yet support for an individual.

Recent advances

Long-read and multi-omics sequencing. Older sequencing produced short fragments that were hard to assemble into complete genomes. Newer long-read platforms (such as Oxford Nanopore and PacBio) read much longer stretches at once, which allows near-complete genomes to be reconstructed straight from a sample and pushes identification down to the strain level. Combined with the move to read RNA and metabolites alongside DNA (multi-omics), the field is shifting from "who is there" toward "what the whole community is doing," across bacteria, fungi, and viruses at once.

The gut virome and phageome. The gut is not only bacteria. It holds a vast population of viruses, most of them bacteriophages that infect the gut bacteria. Improved sequencing is revealing how these phages shape the bacterial community by killing some members and sparing others, adding a whole layer of control we could not see before. This is a young area and much of it is still being mapped.

Computation and AI. Microbiome datasets are enormous, and making sense of them increasingly depends on machine learning to find patterns linking a community to a disease or a treatment response. This is a genuine advance and also a place to stay skeptical, because a pattern found by an algorithm still has to be validated with a mechanism and, ideally, an intervention before it means anything clinically.

The direction of travel is consistent across all three: more complete, more functional, and more able to link a community to what it does in the body. The gap between what we can measure and what we can safely act on is still wide, and closing it is the work of the next decade.

How to Remember

Flora asks "who," microbiome asks "who and what they do." This one line carries the whole conceptual shift. Normal flora came from growing microbes and naming them. The microbiome comes from reading their DNA and asking what their genes do.

The method ladder: who to what. Culture and 16S tell you who is there. Metagenomics tells you who is there and what they could do. Metatranscriptomics and metabolomics tell you what they are actually doing now. Climb the ladder and you gain function; you also pay in cost and difficulty. Picture a factory: 16S counts the workers, metagenomics reads their job descriptions, metatranscriptomics watches who is actually working today, metabolomics measures what came off the line.

Healthy is site-specific: gut wants diversity, vagina wants dominance. A common exam trap is to assume high diversity is always good. In the gut, yes. In the vagina, a healthy community is dominated by one group (Lactobacillus), and diversification is a sign of disease. "Balanced for that site" beats "diverse" as the rule.

Evidence ladder: differs, explains, changes. For any microbiome-and-disease claim, ask three questions in order. Does the microbiome differ in the disease (association)? Can we explain how it contributes (mechanism)? Does changing it change the disease (intervention)? Recurrent C. difficile passes all three. Most conditions pass only one or two. This ladder protects you from headlines.

The C. difficile story is the whole field in one case. Antibiotics remove the community, C. difficile overgrows, restoring the community cures it. Dysbiosis, mechanism, and microbiome-based cure, all in one disease. If you can tell this story, you understand why the field exists.

Key exam facts

Fact	Detail and memory aid
Microbiota vs microbiome	Microbiota = the organisms. Microbiome = the organisms plus their genes and products. "Who" vs "who and what they do."
Why culture missed most of it	Most body microbes will not grow on standard media. Reading DNA directly does not need the organism to grow, so it revealed communities culture never showed.
Scale of the microbiome	An adult carries roughly as many microbial cells as human cells, and far more microbial genes than human genes. Called the "forgotten organ."
16S rRNA sequencing	Reads one marker gene shared by all bacteria. Cheap, fast, bacteria only, usually genus-level, tells you "who" not "what they do."
Shotgun metagenomics	Reads all the DNA. Species and strain level, includes fungi and viruses, and reveals gene function including resistance genes. More costly.
Metatranscriptomics	Reads RNA, so it shows which genes are switched on now, not just which are present. "Could do" (DNA) vs "is doing" (RNA).
Metabolomics	Measures the community's products (short-chain fatty acids, modified bile acids). The most direct read of function.
Dysbiosis	A disease-linked disturbance: loss of diversity, loss of key members, overgrowth of a minor member, or a shift in function. Not simply "bad bacteria appeared."
Site-specific health rule	High diversity is healthy in the gut but not in the vagina, where Lactobacillus dominance is healthy. "Balanced for that site."
Strongest disease link	Recurrent C. difficile: antibiotics remove the community, C. difficile overgrows, restoring the community cures it. Passes association, mechanism, and intervention.
FMT	Transfer of screened donor stool to rebuild the gut community. Established use: recurrent C. difficile. Donor screening is critical; other uses are experimental.
Approved live biotherapeutics	Rebyota (enema, approved Nov 2022) and Vowst (oral capsule, approved Apr 2023), both for preventing recurrent C. difficile. Standardized, regulated medicines derived from donor stool.
Probiotics vs live biotherapeutics	Probiotic effects are strain- and condition-specific with often weak evidence. Approved live biotherapeutics are regulated, trial-tested medicines. Not the same category.
Consumer testing kits	Sequencing may be real, but there is no validated single "healthy" microbiome to compare against. Interpretation runs ahead of the science. Counsel patients with caution.
Evidence ladder	Association (differs) < mechanism (explains how) < intervention (changing it changes the disease). Judge every claim by how far up it climbs.
Contamination trap	Sequencing detects DNA from dead cells and from reagents. In low-biomass sites (airway, the "fetal microbiome" debate), contamination can masquerade as a real community.

Where Students Get Confused

"Microbiome and microbiota are the same word." Close, but exams test the difference. Microbiota is the organisms. Microbiome adds their genes and products. The reason it matters clinically is that two people can carry different species doing the same job, so what the community does can matter more than the exact species list.

"Sequencing found it, so it is really there and alive." Sequencing detects DNA, and DNA persists in dead cells and even in the chemicals used to run the test. A positive sequencing result is evidence of genetic material, not proof of a living, active organism. This matters most in samples with very few microbes, where contamination can look like a real finding. The disputed "fetal microbiome" is the classic example.

"A more diverse microbiome is always healthier." Not everywhere. In the gut, higher diversity generally goes with health. In the vagina, a healthy community is dominated by Lactobacillus, and increasing diversity is a sign of disease, not health. The rule is "balanced for that site," not "diverse."

"The microbiome causes obesity / depression / (name the condition)." Be careful with the word "causes." For most conditions we have an association (the microbiome differs) and sometimes a mechanism (we can explain how it might contribute), but not proof that changing the microbiome changes the disease. Recurrent C. difficile is one of the few where the full chain, including a working treatment, is established. For the rest, "linked to" is more honest than "causes."

"Probiotics and fecal transplants are basically the same idea." They share a goal, supporting a healthy community, but they are very different. A probiotic supplement is usually one or a few strains, effects are strain-specific, and evidence for broad claims is often weak. FMT and the approved live biotherapeutics transfer or reconstruct a whole community and are regulated, trial-tested treatments for a specific condition. Do not equate them.

"16S sequencing tells you what the bacteria are doing." It does not. 16S identifies who is present, usually only to genus level, and only bacteria. To learn what genes the community carries you need shotgun metagenomics; to learn what those genes are actually doing you need metatranscriptomics or metabolomics. Matching the method to the question is the skill being tested.

"A consumer gut-test report showing 'low good bacteria' means I need those supplements." This is the counseling trap. We do not yet have a validated definition of a healthy individual microbiome to compare a person against, results differ between companies and samples, and labeling a specific bacterium "good" or "bad" for one person is a claim the science cannot support. The sequencing can be real while the interpretation is not.

References

• Berg, G., Rybakova, D., Fischer, D., et al. (2020). Microbiome definition re-visited: old concepts and new challenges. Microbiome, 8(1), 103. https://doi.org/10.1186/s40168-020-00875-0
• Gilbert, J. A., Blaser, M. J., Caporaso, J. G., Jansson, J. K., Lynch, S. V., & Knight, R. (2018). Current understanding of the human microbiome. Nature Medicine, 24(4), 392–400. https://doi.org/10.1038/nm.4517
• Cani, P. D. (2018). Human gut microbiome: hopes, threats and promises. Gut, 67(9), 1716–1725. https://doi.org/10.1136/gutjnl-2018-316723
• Knight, R., Vrbanac, A., Taylor, B. C., et al. (2018). Best practices for analysing microbiomes. Nature Reviews Microbiology, 16(7), 410–422. https://doi.org/10.1038/s41579-018-0029-9
• Portincasa, P., Bonfrate, L., Vacca, M., et al. (2022). Gut microbiota and short chain fatty acids: implications in glucose homeostasis. International Journal of Molecular Sciences, 23(3), 1105. https://doi.org/10.3390/ijms23031105
• Cryan, J. F., O'Riordan, K. J., Cowan, C. S. M., et al. (2019). The microbiota-gut-brain axis. Physiological Reviews, 99(4), 1877–2013. https://doi.org/10.1152/physrev.00018.2018
• Feuerstadt, P., Louie, T. J., Lashner, B., et al. (2022). SER-109, an oral microbiome therapy for recurrent Clostridioides difficile infection. New England Journal of Medicine, 386(3), 220–229. https://doi.org/10.1056/NEJMoa2106516
• Khanna, S., Assi, M., Lee, C., et al. (2022). Efficacy and safety of RBX2660 in PUNCH CD3, a phase III, randomized, double-blind, placebo-controlled trial with a Bayesian primary analysis for the prevention of recurrent Clostridioides difficile infection. Drugs, 82(15), 1527–1538. https://doi.org/10.1007/s40265-022-01797-x
• Portnoy, A., et al. (2025). What's new and what's next in fecal microbiota transplantation? A narrative review. (See PMC12377394 for the review summarizing the FDA-approved microbiota-based therapeutics.) https://pmc.ncbi.nlm.nih.gov/articles/PMC12377394/
• Amann, R. I., Ludwig, W., & Schleifer, K. H. (1995). Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiological Reviews, 59(1), 143–169. https://doi.org/10.1128/mr.59.1.143-169.1995
• Portik, D. M., et al. (2024). Unveiling microbial diversity: harnessing long-read sequencing technology. Nature Methods, 21, 954–966. https://doi.org/10.1038/s41592-024-02262-1
• Tille, P. M. (2022). Bailey & Scott's Diagnostic Microbiology (15th ed.). Elsevier.
• Murray, P. R., Rosenthal, K. S., & Pfaller, M. A. (2021). Medical Microbiology (9th ed.). Elsevier.